Researchers from Penn State College of Information Sciences and Technology studied the effectiveness of chatbot based symptom checker apps (CSC) compared to traditional doctor’s visits. It turns out that the apps are not nearly as effective.
People tend to turn to mobile devices to seek medical advice because it is easier and quicker than going to the doctor's office. But often mobile devices cannot give patients the full examination they need.
Penn State researchers found that chatbots are not as effective as in-person doctor visits. Source: Pixabay
Current CSC apps cannot support a full diagnostic process that patients get from a traditional doctor visit. CSC apps can only support five exam processes: establishing patient history, evaluate symptoms, give an initial diagnosis, order more diagnostic tests and give referrals or follow-up treatment recommendations. CSC apps cannot conduct physical exams, provide a final diagnosis and perform or analyze test results.
The team investigated the functionalities of CSC apps with feature reviews, examining user experiences and conducting user interviews. Users reported that CSC apps lack the support they need for comprehensive medical history, flexible symptom input, answer questions and work with diverse diseases or user groups.
Researchers say their findings could inform functional and conversational design updates for healthcare chatbots that could improve their effectiveness. The findings could also help individuals understand the influence of AI and how it could change traditional medical visits.
The team's research will be presented at the American Medical Informatics Association (AMIA) Virtual Annual Symposium in November.
